Glossary
Interview bias
Interview bias is any repeated, predictable error that leads an interviewer to rate an applicant on something other than their ability to do the job.
Everyone who interviews has biases. The goal isn’t to find interviewers with none. It’s to design a process that gives biases less chance to change the outcome.
Common types
| Bias | What it looks like |
|---|---|
| First impression | The first two minutes set the rating, and the rest of the interview is spent confirming it. |
| Halo and horns | One strong point (a famous former employer) or one weak point (a nervous start) affects every other rating. |
| Affinity | Rating people higher because they went to the same school, share a hobby or talk like you. |
| Contrast | A decent applicant looks weak after an excellent one, or strong after a poor one. |
| Confirmation | Asking harder follow-up questions of applicants you already doubt. |
| Central tendency | Giving everyone middle scores to avoid committing either way. |
| Fluency and accent | Mistaking smooth speaking for skill, or treating an accent as a sign of weaker ability. |
Practical ways to reduce it
- Structure the interview. Same questions, same order. See structured interview.
- Use a scoring guide, often called a rubric. Written descriptions of each score make it harder to rate on gut feeling.
- Score before discussing. Each interviewer scores on their own, then the group talks.
- Require evidence. Every rating should point to something the applicant said or did.
- Leave out what doesn’t matter. Photos, school names and hobbies rarely help a first-interview decision.
- Review your results. Where the law allows, compare ratings across groups, for example by sex or ethnicity. Patterns show where bias may be at work. See adverse impact.
An example
Two applicants answer a question about an angry customer. One talks fast and confidently. The other pauses, loses their thread once, then gives a detailed, specific account. Without a scoring guide, the first often gets the higher score. With a guide that asks for a specific situation, clear actions and a result, the second may well score higher, because they covered more of what it asks for.
AI and bias
AI can remove some human biases, like tiredness late in the day or favoring people like you. It can also add new ones if it’s built or instructed badly. The questions to ask are practical: what information does it use, does it show its reasoning, and can people overrule it? LessRounds scores only the transcript, the written record of what was said, not faces or voices. It instructs the AI never to mark anyone down for their language, for switching between languages, or for grammar. No tool, human or AI, should be described as unbiased.